<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->
wav2vec2-base-timit-demo-google-colab
This model is a fine-tuned version of facebook/wav2vec2-base on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.5499
- Wer: 0.3435
Model description
More information needed
Intended uses & limitations
More information needed
Training and evaluation data
More information needed
Training procedure
Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 0.0001
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 1000
- num_epochs: 30
- mixed_precision_training: Native AMP
Training results
Training Loss | Epoch | Step | Validation Loss | Wer |
---|---|---|---|---|
3.599 | 1.0 | 500 | 2.1267 | 0.9976 |
1.016 | 2.01 | 1000 | 0.6193 | 0.5443 |
0.5299 | 3.01 | 1500 | 0.5324 | 0.4889 |
0.3626 | 4.02 | 2000 | 0.4525 | 0.4402 |
0.2854 | 5.02 | 2500 | 0.4266 | 0.4233 |
0.2373 | 6.02 | 3000 | 0.4713 | 0.4082 |
0.1979 | 7.03 | 3500 | 0.4778 | 0.4018 |
0.1761 | 8.03 | 4000 | 0.4585 | 0.3947 |
0.1537 | 9.04 | 4500 | 0.5297 | 0.3946 |
0.1379 | 10.04 | 5000 | 0.4988 | 0.3856 |
0.124 | 11.04 | 5500 | 0.5262 | 0.3852 |
0.11 | 12.05 | 6000 | 0.5545 | 0.3854 |
0.106 | 13.05 | 6500 | 0.5196 | 0.3805 |
0.0918 | 14.06 | 7000 | 0.4515 | 0.3655 |
0.0829 | 15.06 | 7500 | 0.5087 | 0.3722 |
0.0775 | 16.06 | 8000 | 0.4980 | 0.3781 |
0.0685 | 17.07 | 8500 | 0.5564 | 0.3650 |
0.0655 | 18.07 | 9000 | 0.5323 | 0.3672 |
0.0578 | 19.08 | 9500 | 0.5675 | 0.3637 |
0.052 | 20.08 | 10000 | 0.5604 | 0.3664 |
0.0512 | 21.08 | 10500 | 0.5922 | 0.3804 |
0.0431 | 22.09 | 11000 | 0.6379 | 0.3754 |
0.0428 | 23.09 | 11500 | 0.5905 | 0.3764 |
0.0393 | 24.1 | 12000 | 0.5667 | 0.3542 |
0.0326 | 25.1 | 12500 | 0.5612 | 0.3537 |
0.0289 | 26.1 | 13000 | 0.5618 | 0.3475 |
0.0298 | 27.11 | 13500 | 0.5578 | 0.3439 |
0.0264 | 28.11 | 14000 | 0.5547 | 0.3433 |
0.026 | 29.12 | 14500 | 0.5499 | 0.3435 |
Framework versions
- Transformers 4.17.0
- Pytorch 1.12.0+cu113
- Datasets 1.18.3
- Tokenizers 0.12.1